Voice of the Customer (VoC) programs enable business leaders to gather the critical customer feedback they need to address customer concerns and shape the future of their products.
Traditional VoC programs are one-on-one interview based and can generate volumes of unstructured customer interview data. Making that information actionable is where the rub is, according to Maxie Schmidt-Subramanian, principal analyst of customer experience at Forrester.
She cautions us that the issue to be concerned with in VoC is not that the data itself is bad, but the fact that you need the right tools and processes to see the patterns and identify the priorities, and further, you need to put governance in place to address it operationally.
Maxie says that the primary issue with voice of the customer data is that in many organizations it’s not actually driving action.
VoC until recently has tapped into customer needs through focus groups, surveys, and one-to-one interviews. These interviews and surveys can generate some valuable customer insights, but the results are skewed towards customers who are willing to interact in this fashion. This is important to remember, since according to Maxie, only “2-7 percent of customers have an interaction with marketing surveys.”
VoC is about to get a whole lot more effective and efficient, with the augmentation of artificial intelligence and machine learning technology that will enable companies to gather data from all their customers, not just the small sampling that has interacted with your brand directly through a survey.
Maxie tells us that VoC will benefit from AI in three key ways.
First companies will get access to more data and types of data they never had before. Second, they will be able to not only record what is being said by customers, but using AI will be able to analyze pitch and tone to understand a customer’s disposition. And finally Maxie says with AI and machine learning you will be able to mine video at scale and contextually tag them.
“Only 21 percent of people in a CX role communicate metrics in ways that drive action.”—Maxie Schmidt-Subramanian
Voice of the Customer/AI Challenges
According to Maxie, there are some current challenges to be overcome until AI and machine learning can fully work their way into VoC product development efforts.
First, customer experience (CX) professionals need to get more familiar and comfortable with AI overall and understand its technology requirements, processes, benefits and limitations. Since according to Maxie “only 21 percent of people in a CX role communicate metrics in ways that drive action,” it is step number one that CX pros need to get better at translating metrics into action, and leveraging AI is a great way to do that. Another thing Maxie warns us to be wary of is that setting up AI can potentially put a burden on internal broken processes and systems you already have in place.
Maxie tells us that businesses and product development professionals need to keep in mind that AI is not a general purpose solution, but one that needs a significant amount of work and effort to get it implemented and operationalized. You want to ensure you have qualified employees who understand data to do the analysis, feed the data in the right way to the AI and machine learning algorithms, and you need someone to check the data when it comes out to make sure it is right.
But with the right amount of set up and human support, machine learning and AI can have significant benefits in the here and now, as well as into the not-to-distant future. Maxie tells us that machine learning can make CMS systems better, like for forming and updating taxonomies. It can allow your company to predict customer experience scores without talking to customers and to use data to train models to help tailor the customer experience at any point in the journey.
Medallia Touts Power of AI-driven VoC Insights
Krish Mantrapragada, chief product officer, Medallia, an enterprise feedback management and voice of the customer software vendor, says Medallia is already using AI to capture extended VoC feedback from sources far beyond the typical interview or survey.
Leveraging AI and machine learning algorithms and technology, Medallia is mining customer preferences and data from all across the web, through social media, mobile activity, and contact center interactions. They then analyze the data and feedback in real-time and provide actionable workflows to executive, frontline and central office workers.
Through their software Krish says they can measure customer satisfaction, customer loyalty and overall business performance, and provide direction on how to improve them.
“You need to connect with a customer earlier in the journey and predict their behavior.”— Krish Mantrapragada
While Medallia collects VoC direct solicited feedback from customers in the form of surveys, app data, and social media, they also collect unsolicited feedback from customers and non-customers as well, and from the general preferences of the public at large, by examining patterns of behavior through data mining in various public environments (e.g. airports).
Krish tells us that you want to apply machine learning on top of all the direct and unsolicited data you collect and sort it to see if it data points are one-off event or more systemic. Once it is analyzed, Medallia’s AI can spit out a suggested action and can tell you how much impact an action would have on business results.
For example, in a retail environment, there may be very long lines for returns of merchandise, and the store wants to see whether it would make sense to reduce that wait time. Some data that would be collected to help inform that decision would include comments from customers about waiting on long return lines, return process numbers showing returns trending up, and receiving feedback from retail employees.
Looking at purchase data, they would find that people who return things are more inclined to purchase more, so helping expedite this process would lead to more sales. With all these data points, the AI would spit out a suggested action that there should be a separate process for returns than other purchase activities.
NICE inContact Uses AI for Smart Surveys
NICE inContact is a call center and customer experience (CX) platform vendor with more than a decade in the business. Chris Bauserman, VP of product and segment marketing for NICE inContact tells us what you should be thinking about when bringing AI into VoC programs, and how it is changing the playing field now and in the future.
Chris tells us we need to ask ourselves three questions when it comes to data collection and analysis.
- What type of data are you collecting?
- How are you making sense of it?
- How are you applying the data and making it actionable?
One effect AI is having on data is that it is making it more granular, allowing behavioral profiling and contextual customer and channel experiences to scale to levels never seen before. Using AI and machine learning, Chris tells us, his organization is developing what they are calling multi-channel surveys.
As we heard from Maxie earlier about the low customer response rate in surveys, NICE inContact is making them more valuable and relevant by augmenting them with analytics, calls/email/various customer interactions, and VoC inference through indirect feedback. This augmented data gives the ability to cross-reference the data points in a survey in order to get a better profile of your customer needs.
The Future State of VoC + AI
While the use of AI and machine learning in VoC programs have grown by leaps and bounds in the last seven years, there are a lot more exciting changes to expect on the horizon.
Maxie Schmidt-Subramanian of Forrester Research tells us AI will help us get closer to understanding customer emotions by using speech analysis and facial recognition. In this, AI can unlock new types of VoC data we have never seen before.
Another advance to look out for, according to Maxie, is a chat-bot interface for the business data user.
For employees in a company who don’t use data on a regular basis, it can be hard for them to understand what it is telling them and put it into action. A chat-bot can help the interaction of business users to the data by bringing business users closer to the data by making it more conversational; allowing end users to get closer to the data through dashboards; and to enable stakeholders to query data in natural language.
Krish Mantrapragada of Medallia, tells us that future state AI is going to be all about in the moment engagement. This is when companies will be able to engage in a conversation with a customer in real-time, and will be able to hand interactions over to a live person when appropriate.
According to Chris Bauserman, real-time data collection is coming. For example, phone calls and chat discussions can be recorded and then associated together in real-time. He thinks that in the future VoC and AI will be all about going true omnichannel for messaging delivery, instead of the actual multichannel that are going out today.
The Bottom Line
In today's competitive landscape, listening to your customers and integrating their feedback isn't optional, it's a must that can demonstrate measurable improvements and ROI. Combine that with the amount of data organizations are currently collecting and will in the future, and it becomes clear, workers are going to need help making sense of all that data to make it clean and actionable for decision makers. AI is poised to have an impact on a number of industries, perhaps all, but it's clear that voice of the customer will soon feel its effect.